Figure 2.2: Reo calculation - empty bundle Figure 2.3: Reo calculation - full bundle 23
Definition 5 (AMP) An AMP is a unit of currency available for a user, represented by a node, N to spend amplifying visibility and reach of content, through sharing and republishing. 2.2.2 How Reo and AMPs get used in <strong>Synereo</strong> Next, we look at the events <strong>Synereo</strong> users generate and how they are influenced by Reo and AMPs. <strong>With</strong> respect to content sharing, there are essentially two kinds of events, those associated with the production of content and those associated with the consumption of content. We associate production of content with volition, whether it is production of original content or resharing content received. This is in contrast to content becoming more likely to show up in a user’s stream because it has received some form of attention. Conversely, we associate consumption with the deployment of attention. We recognize that there is some overlap in these categories. For example, “liking” a post is an act of volition, but the aim of such an act is to acknowledge deployment of attention. How does Reo figure into the publication of content in the network? Likewise how do AMPs influence the publication process? Let’s consider a user, such as Abed, who wants to share a video and so publishes it to his community of friends. This means that the message with the link to the video is offered to each <strong>Synereo</strong> node associated with one of Abed’s friends in <strong>Synereo</strong>. The node on the receiving end will initiate a process to determine the ordering of that message in Abed’s friend’s stream, and that process makes essential use of Reo and AMPs. Network dynamics and information flow dynamics In order to address this we have to introduce the dynamics of information flow in the network. The most basic and natural intuition is that events flow from node to node over the edges connecting them. As a first approximation we will treat the network as fixed and discuss how content flows in that fixed network. The reality, however, is that the network is dynamic, both at the level of nodes supporting the <strong>Synereo</strong> protocols and at the level of the social network, who is connected to whom, who is friends with whom. Moreover, the network topology, especially the topology of the social network, changes as a direct consequence of various content flows. Abed posts an interesting video. Troy shares it with a friend, acknowledging Abed’s authorship. The friend is impressed and asks Troy for an introduction to Abed. In today’s world, network dynamics and their mutually recursive dependency on information flow dynamics are the bread and butter of social networks. Being able to model this phenomenon succinctly and precisely ultimately requires moving to a different approach, and is another key reason why we introduce a process calculus based model in subsequent sections. The concept of mobility in the π-calculus captures this notion perfectly in a beautiful algebraic theory, rather than in an adhoc manner. We recognize that while the theory of mobile processes is nearly 30 years old, it’s still new to many. So, we get there in stages. Part of the advantage of having a “god’s eye” view is that we can not only see the whole network, in its entirety, but we can see the whole of time, in its entirety. Fortunately, through the magic of recursion, we don’t have to see it 24